Update app.py

app.py

@@ -1,95 +1,176 @@
-# %%bash
+# # %%bash
 
-# # git lfs install
-# # git clone https://huggingface.co/spaces/Xhaheen/meme_world
+# # # git lfs install
+# # # git clone https://huggingface.co/spaces/Xhaheen/meme_world
 
 
-# # pip install -r /content/meme_world/requirements.txt
-# # pip install gradio
-# cd /meme_world
+# # # pip install -r /content/meme_world/requirements.txt
+# # # pip install gradio
+# # cd /meme_world
 
 
-import torch 
-import re 
-import gradio as gr
-from transformers import AutoTokenizer, ViTFeatureExtractor, VisionEncoderDecoderModel 
-import cohere
-import os
-# 
-# os.environ['key_srkian'] = ''
-key_srkian = os.environ["key_srkian"]
-co = cohere.Client(key_srkian)#srkian
-device='cpu'
-encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
-decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
-model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
-feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_checkpoint)
-tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)
-model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)
-
-
-def predict(department,image,max_length=64, num_beams=4):
-  image = image.convert('RGB')
-  image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)
-  clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]
-  caption_ids = model.generate(image, max_length = max_length)[0]
-  caption_text = clean_text(tokenizer.decode(caption_ids))
-  dept=department
-  context= caption_text
-  response = co.generate(
-  model='large',
-  prompt=f'create non offensive one line meme for given department and context\n\ndepartment- data science\ncontext-a man sitting on a bench with a laptop\nmeme- \"I\'m not a data scientist, but I play one on my laptop.\"\n\ndepartment-startup\ncontext-a young boy is smiling while using a laptop\nmeme-\"When your startup gets funded and you can finally afford a new laptop\"\n\ndepartment- {dept}\ncontext-{context}\nmeme-',
-  max_tokens=20,
-  temperature=0.8,
-  k=0,
-  p=0.75,
-  frequency_penalty=0,
-  presence_penalty=0,
-  stop_sequences=["department"],
-  return_likelihoods='NONE')
-  reponse=response.generations[0].text
-  reponse = reponse.replace("department", "")
-  Feedback_SQL="DEPT"+dept+"CAPT"+caption_text+"MAMAY"+reponse
+# import torch 
+# import re 
+# import gradio as gr
+# from transformers import AutoTokenizer, ViTFeatureExtractor, VisionEncoderDecoderModel 
+# import cohere
+# import os
+# # 
+# # os.environ['key_srkian'] = ''
+# key_srkian = os.environ["key_srkian"]
+# co = cohere.Client(key_srkian)#srkian
+# device='cpu'
+# encoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# decoder_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# model_checkpoint = "nlpconnect/vit-gpt2-image-captioning"
+# feature_extractor = ViTFeatureExtractor.from_pretrained(encoder_checkpoint)
+# tokenizer = AutoTokenizer.from_pretrained(decoder_checkpoint)
+# model = VisionEncoderDecoderModel.from_pretrained(model_checkpoint).to(device)
+
+
+# def predict(department,image,max_length=64, num_beams=4):
+#   image = image.convert('RGB')
+#   image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)
+#   clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]
+#   caption_ids = model.generate(image, max_length = max_length)[0]
+#   caption_text = clean_text(tokenizer.decode(caption_ids))
+#   dept=department
+#   context= caption_text
+#   response = co.generate(
+#   model='large',
+#   prompt=f'create non offensive one line meme for given department and context\n\ndepartment- data science\ncontext-a man sitting on a bench with a laptop\nmeme- \"I\'m not a data scientist, but I play one on my laptop.\"\n\ndepartment-startup\ncontext-a young boy is smiling while using a laptop\nmeme-\"When your startup gets funded and you can finally afford a new laptop\"\n\ndepartment- {dept}\ncontext-{context}\nmeme-',
+#   max_tokens=20,
+#   temperature=0.8,
+#   k=0,
+#   p=0.75,
+#   frequency_penalty=0,
+#   presence_penalty=0,
+#   stop_sequences=["department"],
+#   return_likelihoods='NONE')
+#   reponse=response.generations[0].text
+#   reponse = reponse.replace("department", "")
+#   Feedback_SQL="DEPT"+dept+"CAPT"+caption_text+"MAMAY"+reponse
   
      
-  return reponse 
+#   return reponse 
 
 
 
-# input = gr.inputs.Image(label="Upload your Image", type = 'pil', optional=True)
+# # input = gr.inputs.Image(label="Upload your Image", type = 'pil', optional=True)
  
  
  
-output = gr.outputs.Textbox(type="text",label="Meme")
-#examples = [f"example{i}.jpg" for i in range(1,7)]
-#examples = os.listdir() 
-examples = [f"example{i}.png" for i in range(1,7)]
+# output = gr.outputs.Textbox(type="text",label="Meme")
+# #examples = [f"example{i}.jpg" for i in range(1,7)]
+# #examples = os.listdir() 
+# examples = [f"example{i}.png" for i in range(1,7)]
 
-#examples=os.listdir()
-#for fichier in examples:
-  #  if not(fichier.endswith(".png")):
-   #     examples.remove(fichier)
+# #examples=os.listdir()
+# #for fichier in examples:
+#   #  if not(fichier.endswith(".png")):
+#    #     examples.remove(fichier)
         
-description= " Looking for a fun and easy way to generate memes? Look no further than Meme world! Leveraging large language models like GPT-3PT-3  / Ai21 / Cohere, you can create memes that are sure to be a hit with your friends or network. Created with ♥️ by Arsalan @[Xaheen](https://www.linkedin.com/in/sallu-mandya/). kindly share your thoughts in discussion session and use the app responsibly #NO_Offense \n \n built with ❤️   @[Xhaheen](https://www.linkedin.com/in/sallu-mandya/)"
-title = "Meme world 🖼️"
-dropdown=["data science", "product management","marketing","startup" ,"agile","crypto" , "SEO" ]
+# description= " Looking for a fun and easy way to generate memes? Look no further than Meme world! Leveraging large language models like GPT-3PT-3  / Ai21 / Cohere, you can create memes that are sure to be a hit with your friends or network. Created with ♥️ by Arsalan @[Xaheen](https://www.linkedin.com/in/sallu-mandya/). kindly share your thoughts in discussion session and use the app responsibly #NO_Offense \n \n built with ❤️   @[Xhaheen](https://www.linkedin.com/in/sallu-mandya/)"
+# title = "Meme world 🖼️"
+# dropdown=["data science", "product management","marketing","startup" ,"agile","crypto" , "SEO" ]
 
-article = "Created By :  Xaheen "
+# article = "Created By :  Xaheen "
 
-interface = gr.Interface(
-        fn=predict,
-        inputs = [gr.inputs.Dropdown(dropdown),gr.inputs.Image(label="Upload your Image", type = 'pil', optional=True)],
+# interface = gr.Interface(
+#         fn=predict,
+#         inputs = [gr.inputs.Dropdown(dropdown),gr.inputs.Image(label="Upload your Image", type = 'pil', optional=True)],
         
-        theme="grass",
-        outputs=output,
-        examples =[['data science', 'example5.png'],
-            ['product management', 'example2.png'],
-            ['startup', 'example3.png'],
-            ['marketing', 'example4.png'],
-            ['agile', 'example1.png'],
-            ['crypto', 'example6.png']],
-        title=title,
-        description=description,
-        article = article,
-    )
-interface.launch(debug=True)
+#         theme="grass",
+#         outputs=output,
+#         examples =[['data science', 'example5.png'],
+#             ['product management', 'example2.png'],
+#             ['startup', 'example3.png'],
+#             ['marketing', 'example4.png'],
+#             ['agile', 'example1.png'],
+#             ['crypto', 'example6.png']],
+#         title=title,
+#         description=description,
+#         article = article,
+#     )
+# interface.launch(debug=True)
+
+
+
+
+
+
+
+
+
+
+
+# Step 2: Set up the Gradio interface and import necessary packages
+import gradio as gr
+import openai
+from transformers import VisionEncoderDecoderModel, ViTImageProcessor, AutoTokenizer
+import torch
+from PIL import Image
+
+# Step 3: Load the provided image captioning model
+model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+feature_extractor = ViTImageProcessor.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+
+# Step 4: Create a function to generate captions from images
+max_length = 16
+num_beams = 4
+gen_kwargs = {"max_length": max_length, "num_beams": num_beams}
+
+def generate_caption(image):
+    image = Image.fromarray(image.astype('uint8'), 'RGB')
+    if image.mode != "RGB":
+        image = image.convert(mode="RGB")
+    pixel_values = feature_extractor(images=[image], return_tensors="pt").pixel_values
+    pixel_values = pixel_values.to(device)
+    output_ids = model.generate(pixel_values, **gen_kwargs)
+    caption = tokenizer.decode(output_ids[0], skip_special_tokens=True).strip()
+    return caption
+
+
+# Step 5: Create a function to generate memes using the GPT-3 API
+def generate_meme(caption, department):
+    openai.api_key = os.environ["key"]
+    prompt = f"Create a non-offensive meme caption for the following image description in the context of {department} department: {caption}"
+    response = openai.Completion.create(engine="text-davinci-002", prompt=prompt, max_tokens=50, n=1, stop=None, temperature=0.7)
+    meme_caption = response.choices[0].text.strip()
+    return meme_caption
+
+# Step 6: Define the main meme generation function
+def meme_generator(image, department):
+    caption = generate_caption(image)
+    meme_caption = generate_meme(caption, department)
+    return meme_caption
+
+examples = [f"example{i}.png" for i in range(1,7)]
+
+# Step 7: Launch the Gradio application
+image_input = gr.inputs.Image()
+department_input = gr.inputs.Dropdown(choices=["data science", "product management","marketing","startup" ,"agile","crypto" , "SEO" ])
+output_text = gr.outputs.Textbox()
+
+gr.Interface(fn=meme_generator, inputs=[image_input, department_input], outputs=output_text, title="Meme world!",description= " Looking for a fun and easy way to generate memes? Look no further than Meme world! Leveraging large language models like GPT-3PT-3  / Ai21 / Cohere, you can create memes that are sure to be a hit with your friends or network. Created with ♥️ by Arsalan @[Xaheen](https://www.linkedin.com/in/sallu-mandya/). kindly share your thoughts in discussion session and use the app responsibly #NO_Offense \n \n built with ❤️   @[Xhaheen](https://www.linkedin.com/in/sallu-mandya/)",        theme="grass",
+    
+        examples =[['example5.png','data science' ],
+            ['example2.png','product management'],
+            ['example3.png','startup'],
+            ['example4.png','marketing'],
+            ['example1.png','agile'],
+            ['example6.png','crypto']]).launch(debug=True)
+
+
+
+
+
+
+
+
+
+
+